What is Uptime?

Uptime, explained

Uptime is a core reliability metric, often written into service level agreements. Availability targets are commonly described in nines. For example, 99.9 percent uptime, three nines, allows roughly 8.8 hours of downtime per year, while 99.99 percent, four nines, allows under an hour.

Achieving high uptime requires redundancy across servers and data centers, load balancing, automated failover, careful deployment practices, and good monitoring and observability so problems are detected quickly. Planned maintenance may or may not count against uptime, depending on the agreement. Each additional nine is usually much harder and more expensive to achieve, so teams choose targets that match how critical a service is. Site reliability engineering often uses error budgets to balance reliability with shipping new features.